Freeze-drying of Proteins

5/21/2018 Freeze-drying of Proteins

1/13

Biotechnol. Appl. Biochem. (2004)39, 165177 (Printed in Great Britain) 1

REVIEW

Freeze-drying of proteins: some emerging concerns

Ipsita Roy and Munishwar Nath Gupta1

Chemistry Department, Indian Institute of Technology, Delhi, Hauz Khas, New Delhi 110016, India

Freeze-drying (lyophilization) removes water from

a frozen sample by sublimation and desorption. It

can be viewed as a three-step process consisting

of freezing, primary drying and secondary drying.

While cryoprotectants can protect the protein from

denaturation during early stages, lyoprotectants are

needed to prevent protein inactivation during drying.The structural changes as a result of freeze-dry-

ing have been investigated, especially by FTIR (Fourier-

transform IR) spectroscopy. In general, drying results

in a decrease of -helix and random structure and

an increase in -sheet structure. In the case of basic

fibroblast growth factor and -interferon, enhanced

FTIR showed large conformational changes and

aggregation during freeze-drying, which could be

prevented by using sucrose as a lyoprotectant. It is now

well established that structural changes during freeze-

drying are responsible for low activity of freeze-dried

powders in nearly anhydrous media. Strategies such

as salt activation can give activated enzyme powders,

e.g. salt-activated thermolysin-catalysed regioselective

acylation of taxol to give a more soluble derivative for

therapeutic use. In the presence of moisture, freeze-

dried proteins can undergo disulphide interchange

and other reactions which lead to inactivation. Such

molecular changes during storage have been described

for human insulin, tetanus toxoid and interleukin-2.

Some successful preventive strategies in these cases

have also been mentioned as illustrations. Finally, it

is emphasized that freeze-drying is not an innocuous

process and needs to be understood and used carefully.

Introduction

Freeze-drying (lyophilization) is perceived to be a gentle

process for concentrating or drying biologically active

substances. Enzymologists have been using it as a routine

technique for more than four decades, and pharmaceutical

industries also adopted it quite some time ago. This is

because enzyme/protein- or peptide-based drugs are more

stable in solid (as compared with solution) form. Thus

freeze-dried powders offer advantages at the storage and

shipping/distribution stages. They also increase the shelf-life

at the end-user stage [1]. The two reasons why a greater

understanding of freeze-drying has assumed an urgent and

considerable relevance to life science and pharmaceutical

science are as follows: (1) it has been realized in recent

years that stability of freeze-dried powdersdepends critically

on how the freeze-drying is carried out [2,3]; this isof great concern to producers of enzymes and other

biologically active proteins; (2) in the last few years it

has been realized that the low activity of freeze-dried

enzyme powders in nearly anhydrous organic solvents is

largely due to protein inactivation at the stage of freeze-

drying; non-aqueous enzymology is employed for many

biotechnologically important bioconversions, synthesis and

resolution of stereoisomers [46].

In view of these factors, there is a need to evolve

techniques to minimize this form of inactivation and improve

catalytic efficiency during the target process. This, of course,

involves basic understanding of freeze-drying, which is a

complex process.

Basic process [7]

Freeze-drying removes water from a frozen sample by

sublimation and desorption. All freeze-driers (lyophilizers),

whether a simple glass unit fabricated in a glass-blowing

laboratory or a sophisticated industrial-scale unit, have

two essential design features: a low-pressure chamber to

which the frozen sample is attached and a cold finger

or cold trap which collects the sublimated or desorbed

ice. The process of freeze-drying can be visualized interms of three steps [2]. These three stages have been

discussed in detail fairly recently [8], so it may thus suffice

here to present a brief qualitative picture only. It may,

Key words: cryoprotectants, freeze-drying, lyoprotectants, protein

aggregation, protein lyophilization, storage stability of pharmaceuticals.

Abbreviations used: FTIR, Fourier-transform IR; PMR, proton magnetic

resonance;Tg, glass transition temperature of the dried product; Tg , glass

transition temper ature associated with maximum freeze concentration;

Tmc , critical temperature of molecular mobility;Wg , amount of frozen

water at glass transition temperature; pCl, log [Cl].1 To whom correspondence should be addressed (e-mail

mn [email protected]).

C2004 Portland Press Ltd


2/13

166 I. Roy and M. N. Gupta

however, be stressed that process validation requires strict

control of various parameters at all the three stages dur-

ing various cycles (http:/ www.fda.gov/ ora/ inspect ref/igs/

lyophi.html).

(a) Initial freezing

Freezing involves formation of ice nuclei. Depending upon

many factors, especially the cooling rate, supercooling may

occur. Thus ice formation may, in fact, take place well

below 0 C [8]. Faster freezing rates give rise to small ice

crystals. The result is faster sublimation at the primary

drying phase, but slower secondary drying. Removal of

water by freezing increases the solute concentration and

hence the viscosity. At some stage, a saturation value is

reached and no further increase in concentration/viscosity

is possible. At this stage, the glass transition [involving the

parameter Tg

(glass transition temperature associated withmaximum freeze concentration)] takes place. A glass is a

supersaturated, thermodynamically unstable liquid with the

viscosity as high as 101214 Pa s [9]. In other words, a glass

transition is not a true thermodynamic change in state, but

instead is a kinetic limit at high viscosity [10]. Two other

parameters that are important, areWg (the unfrozen water

at this temperature) andTg(the glass transition temperature

of the dried product) [11]. Both Wg andTgcan be measured

by differential scanning calorimetry [11].

(b) Primary drying

This is the stage at which ice separated from the solute

phase is removed by sublimation. If the sample is thick, mass

transfer constraints decrease the rate of sublimation [8]. The

temperature of the sample is a rather critical parameter; if

the temperature rises aboveTg, the ice would melt into the

solute phase. At the same time, a temperature rise of 1 C is

reported to make drying faster by 13 % [8]. Hatley [11] has

shown that carrying out primary drying aboveTg produces

freeze-dried samples of inferior quality.

(c) Secondary drying

This stage begins after all the frozen water has sublimed

and thus can be facilitated by increasing the product

temperature. In the case of proteins, this temperature hasto be chosen while keeping in view the thermal stability of

proteins. The bound water (to protein molecules) or the

water molecules trapped in the glass phase are removed

during this stage. Up to the 2 % level, the water is re-

moved quickly, and this process slows down thereafter. This

is because, as the sample dries, diffusion of water molecules

through the sample becomes more difficult. It is a function

of the porosity of the sample and does not depend much

upon the sample thickness [8].

In the case of food proteins, there is extensive data

to indicate that storage temperature of the freeze-dried

proteins should be kept higher than their Tg. This is

necessary to prevent collapse. The volumetric shrinkage

after collapse, the importance of sample mass, etc., have

also been discussed. There are sufficient data to suggest

that similar factors are important in the case of freeze-

dried proteins [8]. Excipients, which result in higher Tgor higher collapse temperature, result in greater stability

of freeze-dried proteins and vice versa. It is also known

that excipients that facilitate phase separation during freeze-

drying, accelerate protein inactivation [12].

The presence of excipients during freeze-drying

obviously leads to a more complicated phase diagram at

the freezing and primary drying stages [13]. During the

subsequent discussion of the effect of excipients, the ex-

planations have relied on the more widely accepted theory

of water substitution [8]. However, it may be worth

mentioning straight away that the vitrification hypothesispresents another viewpoint. According to some workers,

proteins in glassy state exhibit decreased molecular motion,

which slows down many degradative processes [8,9].

However, as Craig et al. [8] have pointed out, a clear link

between Tg of the formulation and storage stability of the

product has not been established.

Bound water and protein activity

The classical experiments by Rupley et al. [14] indi-

cated the importance of a minimum amount of water in

making the enzyme molecule catalytically active. Using heatcapacity and IR spectroscopy, Rupley et al. [14] showed that

water molecules were deposited on dry lysozyme and the

latter became biologically active even before the monolayer

of water molecules around the enzyme molecule had

formed. Subsequent work indicated that water deposition

follows a definite sequence. First it gets deposited on

charged and polar amino acids, and, next, around the hy-

drophobic clusters. The former stage is vital, since, in the

absence of water molecules, the side chains of amino acids

interact with each other and lock up the conformation.

This robs the enzyme molecule of the flexibility that is

necessary for its catalytic activity [15,16]. This rigidity, inthe absence of enough water, is in fact responsible for many

enzyme molecules being stable in the dry state even at

100 C in anhydrous organic solvents [17], even though the

same enzyme may get completely inactivated at much lower

temperatures when heated in its aqueous solution.

Freeze-drying and protein stability

There are two kinds of protein stability which are

of importance. First is the stability during storage and

transport of the protein. This is important to both enzyme



3/13

Freeze-dried proteins 1

manufacturers, who have to ship the biologically active

proteins to distant places, and to the life scientist, who

has to decide storage conditions and permissible shelf-life.

It is of special concern in the case of pharmaceutically

important proteins and vaccines in the context of healthcare

in developing countries where a cold chain is not necessarily

available for storage. The second kind of stability is

operational stability, which describes how stable the enzyme

is during catalysis. This stability can often be higher, since the

presence of the substrate may enhance the conformational

stability of the enzyme [18]. The operational stability can

also be drastically reduced if the components of the

substrate preparation include a molecule which inactivates

the enzyme. A good illustration of this is the reduced

operational stability of lactase during hydrolysis of whey

lactose [19]. In the following discussion, we are mostly

concerned with storage stability.It has been known for a long time that most freeze-

dried RNase A samples contain dimers and aggregates [20].

The amount of these enzymically active aggregates increases

significantly if the freeze-drying is carried out from 50 %

acetic acid. An interesting result worth recalling is that a

dimer formed from inactive alkylated monomers modified

at His12 and His119 had 50 % activity, since these modi-

fied monomer molecules fashioned an active site consisting

of unmodified His12 and His119 [21]. The second important,

even if obvious, comment is that the mechanisms of protein

inactivation are not different, irrespective of which stress

factor is applied. Proteins at high temperatures in aqueous

buffers or in anhydrous organic solvents or at high pressure,

follow a similar mechanistic route (at the molecular

level) during their inactivation. Protein inactivation can be

either reversible or irreversible [22]. During the reversible

phase, the protein chain unfolds and can go back to

the native structure if the denaturing factor is removed.

Exposure to moderately high temperatures such as 40

45 C for a few minutes or 8 M urea are two examples

of such denaturing conditions. Prolonged heat treatment

at higher temperatures will cause irreversible changes in

the protein molecule. Molecular mechanisms for these

irreversible changes are fairly well understood and have

been briefly discussed [23]. Apart from conformationalchanges (which may be reversible at early stages), proteins

eventually undergo a number of other structural changes

[13]. Unfolding of protein chains exposes many hydrophobic

residues which were buried in the native structure. Inter-

molecular hydrophobic interaction between such residues

generally leads to protein aggregation. The disulphide ex-

change reaction, another inactivation mechanism, causes

mispairing of thiol groups, either of intermolecular or intra-

molecular nature. Inactivation mechanisms also include

some chemical changes which lead to irreversible loss of

activity in solution or during the post-freeze-drying stage.

Such changes can also occur, especially if extreme pH

conditions prevail during freeze-drying. Cases of hydro-

lysis of peptide bonds (next to aspartic acid residues),

deamidation of asparagine residues and glutamine, and

racemization of amino acid residues have been well-

documented [24]. In the case of glycoproteins or if free

reducing sugars are present in the sample, the Maillard

reaction plays an important role in determining the overall

stability [24]. Many of these reactions can be minimized

by ensuring that the protein is not exposed to high temp-

eratures at any stage. The relative role played by a particular

chemical change may depend considerably upon the stress

factor or the particular protein. The third point, which

is also worth remembering, is that (although the same

or similar inactivation mechanisms operate), structural

changes during freeze-drying itself and the post-freeze-

drying stage (storage) need to be identified. This distinctionis often overlooked. This happens because, for a variety

of end applications, one tends to look at the activity at

the time of usage. For example, in the pharmaceutical

industry, all operations beyond purification stage, such as

freezing, thawing, formulation, sterile filtration, filling, freeze-

drying and inspection are called formulation or fill-finish

operations [25]. The whole issue is complicated further

by the fact that reconstitution conditions may decide the

ultimate recovery. Webb et al. [26], while working with

recombinant human interferon-, showed the potential

for recovery of native protein using the appropriate

reconstitution conditions, even though the protein is non-

native in the lyophilized state.

Protein stability in solution [23,27]

During early stages of freeze-drying, the protein is in

fact in solution. Partial unfolding, which happens during

this interfacial phenomenon, is largely reversible. Millqvist-

Fureby et al. [28] have described the use of electron

spectroscopy for chemical analysis to show the existence

of surface of freeze-dried solid enriched in protein. Earlier,

Eckhardt et al. [29] showed that rapid freezing promotes

aggregation of human growth hormone. This is in agreement

with the observation of Hsu et al. [30] that the rateof development of turbidity in reconstituted freeze-dried

solids increased with the increased surface areas of the sol-

ids. In fact, Sarciaux et al. [31] have shown that annealing

reduces the specific surface area of freeze-dried solids and

aggregation of bovine IgG.

Protein stability during freeze-thawing

Arakawa et al. [2] have provided a comprehensive

treatment of factors affecting protein stability during freeze-

drying. Their review makes a distinction between the



4/13


Figure 1 Effect of freezing rate on recovery of activity of lactatedehydrogenase

Reproduced from [32] with the permission of Wiley Periodicals,

Inc., a Wiley Company. c 2003 (http://www3.interscience.wiley.com/cgi-

bin/jabout/71002188/ProductInformation.html).

freezethawing and freeze-drying processes in the context

of protein stability. Figure 1 shows the importance of freezing

and thawing rates on the recovery of enzyme activity.

The maximum recovery of activity occurred with slow

freezing and fast thawing. It is reported that slow freezing

leads to large ice crystals, whereas fast freezing leads to

fine ice crystals [32]. The latter increased the interfacial

area considerably and led to denaturation. Figure 2 shows

that freezing damage increased with the addition of salt

throughout the range of freezing rates. This is attributed

to the formation of Na2HPO4 from K3PO4 and NaCl;

Na2HPO4 precipitates at low temperature and reduces

the pH of the unfrozen buffer. Arakawa et al. [2] have

presented experimental evidence to show that a co-solute

has the same stabilization or destabilization effect during

freezethawing as it has during the solution stage. However,

the co-solutes behave differently during freeze-drying,

which led Crowe et al. [33] to conclude that mechanisms

of destabilization during freeze-drying are different

from freezethawing. The osmolytes protect proteinsduring freezethawing and the preferential exclusion theory

developed by Timasheff and co-workers [34,35] can be

extended to freezethawing. According to the preferential

exclusion theory, the solutes are excluded from the vicinity

of the protein and the protein is preferentially hydrated.

Numerous compounds (cryoprotectants), with widely

differing chemical structures, such as sugars, amino acids,

amines, polyols and some salts, if present with the protein,

act as protectors during freezethawing. Compounds such

as MgCl2, guanidinium chloride, KSCN and urea show weak

preferential exclusion or preferential binding and destabilize

Figure 2 Effect of freezing rate on the recoveries of activities of (a) lactate

dehydrogenase and (b) alcohol dehydrogenase in the presence of NaCl

Reproduced from [32] with the permission of Wiley Periodicals,

Inc., a Wiley Company. c 2003 (http://www3.interscience.wiley.com/cgi-

bin/jabout/71002188/ProductInformation.html).

the protein in solution, and thus are expected to play the

same role during freezethawing. Similarly, pH, protein

concentration, and the presence of cofactors and allosteric

modifiers, which favour protein stability in solution, will

also stabilize a protein during freezethawing. This is not

unexpected, since during freezing, protein molecules are

present in the non-ice phase along with the co-solute. The

concentration of the co-solute is also very high, since bulk

water has formed ice.

Lyoprotectants: protein stabilityduring freeze-drying

Freeze-drying involves two denaturing conditions: freezing

and drying. Stabilization of proteins during freeze-drying

can be attempted by adding excipients to the protein

solution. In order to be effective, these excipients have to

protect the protein structure during both stress factors

of freezing and drying. As pointed out by Arakawa et al.

[2], Many effective cryoprotectants have no stabilizing

effect during dehydration. The only solutes that do provide

stabilization are disaccharides. This mechanism is not yet



5/13


fully understood but it can not involve preferential exclusion

of co-solutes. Carpenter and Crowe [36,37] have used

FTIR (Fourier-transform IR) spectroscopy to show that

certain carbohydrates protect proteins by H-bonding, a role

fulfilled by water molecules which are taken away during

drying. However, in the case of lysozyme, FTIR spectroscopy

showed that the presence of trehalose during freeze-drying

did not prevent the structural changes completely [38]. The

work of Lippert and Galinski [39] with phosphofructokinase

and lactate dehydrogenase also showed H-bonding to be

responsible for stabilization during the drying process.

Prestrelski et al. [40] used resolution-enhanced FTIR

to show that basic fibroblast growth factor, interferon-

and -lactalbumin underwent large conformational

changes and aggregation during freeze-drying, and demon-

strated that these changes could be prevented by the

presence of sucrose. Other carbohydrates such as glucosefail to do so, since they fail to act as cryoprotectants during

freezing. The concentration at which glucose could have

worked as a cryoprotectant leads to crystallization of the

sugar during freeze-drying, so that the molecule is no

more available for H-bonding with the protein. Carpenter

et al. [41] and Prestrelski et al. [42] have suggested a

two-component excipient system in which poly(ethylene

glycol) acts as a cryoprotectant and carbohydrates act as

lyoprotectants. In such systems, glucose at low concen-

trations does act as a lyoprotectant. Prestrelski et al.

[42] have also shown that, unless the native structure is

preserved during the drying process (with the help of

excipients), rehydration does not lead to the recovery

of active proteins. However, later work by Webb et al. [26]

raises the possibility of activity recovery from non-active

structure at the reconstitution stage.

Strambini and Gabellieri [43] tracked the amide I signal

(by FTIR) and the secondary structure after rehydration

(monitored by CD spectroscopy) and biological activity of

actin (measured by polymerization assay). It was found that:

(i) there was loss of secondary structure during freeze-

drying; (ii) structural perturbations induced by dehydration

are still retained, up to some degree, upon rehydration;

(iii) unfolding induced by freeze-drying actin with 1 %

sucrose appeared to be fully reversible upon rehydration.The CD structure of the rehydrated sample is almost

identical to that of native actin; (iv) however, the level

of recovered activity (33 %) in the above case was low.

Thus monitoring of secondary structure in the solid state

and rehydrated samples could be misleading if the data

are directly interpreted in terms of recovery of biological

activity. The authors observe, Protein activity depends

upon higher order structural organization in addition to

secondary structure. Another important observation made

by these authors is that the presence of co-solutes at

the rehydration stage may also promote refolding over

aggregation. Carrasquillo et al. [44] found no correlation

between the increased conformational stability in aqueous

solution and the freeze-drying-induced structural changes

in -chymotrypsin. This is contradictory to the earlier

observation of Arakawa et al. [2] that, in general, any factor

that alters protein stability in non-frozen aqueous solution

will tend to have the same qualitative effect during freeze-

thawing. A useful observation made is that -sheet proteins

seem to resist such changes better than -helical proteins.

Another practical observation is by Jiang and Nail

[45] who worked with catalase, -galactosidase and lactate

dehydrogenase, and concluded that the most important

drying process variable affecting recovery of biological

activity was residual moisture level, with a considerable

decrease in recoverable activity being associated with

moisture levels below 10 %.

The freeze-drying of multimeric proteins has also beeninvestigated. In addition to the prevention of unfolding,

preferentially excluded solids also promote quaternary-

structure formation. Thus, in the case of lactate dehydro-

genase, maintenance of the tetrameric form in the frozen

state during primary drying was found to be critical for

enzyme activity after freeze-drying, and thus BSA and

polyvinylpyrrolidone were found to be useful additives [46].

Yoshioka et al. [47] have examined the NMR-relaxa-

tion-based critical temperature of molecular mobility (Tmc)

and Tg (by differential scanning calorimetry) of freeze-

dried -globulin formulations containing different polymer

excipients. The molecular mobility of water in the

formulations was measured by PMR (proton magnetic

resonance) and dielectric relaxation spectrometry. Not

surprisingly, Tmc was found to increase as the extent

of bound water increased. As Tmc reflected the stability

of formulations, dextran was found to be a better additive

as compared with methylcellulose.

Freeze-dried enzyme powders in

nearly anhydrous media

Most enzymology is concerned with reactions of enzymes

in aqueous buffers. Even in the in vivo context, aqueousmedium is the one that is relevant for discussing the

biological activity of enzymes and proteins. However, use of

enzymes in non-aqueous media offers numerous advantages

(Table 1), especially in the synthesis of fine chemicals

(including drug intermediates). Freeze-dried enzymes have

often been used in such applications [5]. Thus it is useful

to discuss the behaviour of freeze-dried enzyme powders

in such media. It was in the late 1960s that Price and co-

workers used chymotrypsin and xanthine oxidase in dry

organic solvents and found these enzyme powders to be

catalytically active [4850]. However, it was in the 1980s



6/13


Table 1 Advantages of using enzymes in non-aqueous media [4,5]

Hydrolases can be used for synthesis

Greater storage stability

No contamination, because microbial growth is absent

A large number of media are possible, with a wide range ofpolarity and hydrophobicity

Possible range of reaction temperature is expanded

Increased repertoire of reactions, in terms of regiospecificity,

stereospecificity, involvement of compounds soluble in organic

media, etc.

Biotransformations by whole cells are also possible

It is possible to control stereoselectivity by varying temperature

and water activity

Separation of reactants and products from the enzyme is not a

problem, since the enzymes are practically insoluble in organic

solvents

that modern phase of non-aqueous enzymology started,

with some pioneering work from various groups [5155].It was shown that: (i) freeze-dried powders of several

enzymes, such as -chymotrypsin, subtilisin, lipases and

tyrosinase are active in various organic solvents [51,53,54].

Such media have been described as nearly anhydrous,

low-water-containing [56] or neat organic solvents [57];

(ii) enzyme powders gave maximum activity when freeze-

dried from an aqueous buffer with a pH equivalent to the

optimum pH of the enzyme. This process was called pH

tuning and the phenomenon called pH memory (it was

as if the enzyme remembered the pH of the solution from

which it had been lyophilized [58]). This phenomenon will

be described in greater detail later on; (iii) in general, more

hydrophobic solvents, when used as reaction media, gave

higher catalytic rates as compared with hydrophilic solvents.

For example, subtilisin showed a Vmax/Km ratio of 2

103 min1 in n-octane as compared with 4 are the most suitable

and the solvents in the intermediate range (between 2 and 4)

are unpredictable [61]. The general model, which is accepted

in this regard, is that the more hydrophilic solvents strip off

the essential water layer from the enzyme and render the

protein molecules inactive [53,62]. This happens because, in

the absence of water, the side chains of amino acids start

interacting with each other and the conformation loses the

necessary flexibility for catalysis.

An interesting development has been that the essential

water layer can be replaced by polar solvents [63]. The

solvents having log P values between 1.08 and 1.93 have

been recommended for this purpose [64].

Low catalytic activity of freeze-dried

enzyme powders

Klibanov [65] and Lee and Dordick [66] have discussed

the reasons responsible for the low activity of freeze-

dried enzyme powders in organic solvents. Diffusional

limitations, unfavourable energetics of substrate desolvation,

transition-state destabilization and conformational changes

during freeze-drying all contribute to various extents. A

clear quantitative picture about their relative roles with an

adequate number of enzymes is not available. In fact, words

like dramatic, etc., so frequently used in non-aqueous

enzymology, have succeeded in taking attention away from

the fact that enzyme powders do have low catalytic power in

organic solvents as compared with aqueous buffers. There

are only limited data on the relative catalytic efficiencies in

the two different media. The kcat values for esterase and

transesterification activities of subtilisin (with hexane as the

reaction medium) are 5.9 103 and 1.04 101 M1 s1

respectively [66]. Two lines of investigation, namely spec-

troscopic studies (mostly FTIR spectroscopy) of freeze-

dried enzyme powders and partially successful attempts

at preventing loss of enzyme activity during freeze-drying,

indicate that freeze-drying is the key factor responsible forthe low activity of enzyme powders in organic media.

Activated freeze-dried enzymes for

use in nearly anhydrous media

Lee and Dordick [66] have recently provided an excellent

list of strategies which yield freeze-dried enzyme powders

showing high activity in nearly anhydrous media. In most of

these strategies, the common feature is to use an additive

which prevents drastic structural changes in the enzyme

during drying. An important point made by the authors

is . . . because enzyme activity and selectivity are kineticallyequivalent (through the catalytic efficiency term, Vmax/Km),

many methods discovered in recent years have also led

to tailored enzyme selectivity [66]. Table 2 lists some

examples of enzymes that have exhibited the phenomenon

of activation in organic solvents.

pH-tuned freeze-dried enzymes

pH tuning is necessary in many cases to obtain significant

activity in non-aqueous media. Later work, especially by

Halling and co-workers, has given fairly good understanding



7/13


Table 2 List of additives used to activate enzymes in non-aqueous media

Enzyme Solvent Additive Activation factor Reference(s)

Protease (Aspergillus or yzae) Anhydrous pyridine Sorbitol 64 [116]

Sucrose 38 [116]Xylitol 32 [116]

Trehalose 26 [116]

Mannitol 20 [116]

Poly(ethylene glycol) 8 [116]

-Chymotrypsin Anhydrous pyridine Sorbitol 19 [116]

Subtilisin Carlsberg Anhydrous pyridine Sorbitol 45 [116]

Anhydrous 17--Oestradiol 10 [82]

2-methylbutan-2-ola

Hexane KCl (98 %, v/v) 1920 [117]

Lipase

Pseudomonas cepacia Anhydrous pyridine Sorbitol 3 [118]

Mucor javanicus Toluene KCl (98 %, v/v) 5 [119]

Pseudomonas stutzeri( lipase TL) Anhydrous 2-methylbutan-2-ol 17--Oestradiol 3-carboxymethyl ether 26 [82]

Humicola lanuginosa Hexane KCl (98 %, v/v) 46 [120]

Humicola lanuginosa Hexane 18-Crown-6c 40 [74]

Mucor javanicus Hexane KCl (98 %, v/v) 11 [74]

Lipoprotein lipase (Pseudomonassp.) Anhydrous pyridine Sorbitol 8 [116]

Horseradish peroxidase Acetone (97 %, v/v) o-Hydroxybenzyl alcohol 62 [118]

Propan-2-olb (99.4 %, v/v) o-Hydroxybenzyl alcohol 65 [118]

Acetone (97 %, v/v) Urea (8 M) 56 [79]

Water-saturated hexane Urea (8 M) 350 [79]

Soybean peroxidase Acetone (99.5 %, v/v) o-Hydroxybenzyl alcohol 400 [118]

m-Hydroxybenzyl alcohol 100 [118]

trans-1,2-Cyclohexanediol 110 [118]

Guaiacol+ poly(ethylene glycol) 450 [118]

trans-1,2-Cyclohexanediol 800 [118]

+ poly(ethylene glycol)

Phospholipase D Chloroform KCl (98 %, v/v) 10 [121]

Penicillin amidase Hexane KCl (98 %, v/v) 750 [122]

Acetonitrile KCl (98 %, v/v) 225 [122]a Synonym t-amyl alcohol.b Synonym isopropanol.c A crown ether.

of the phenomenon of pH memory [67]. It is now clear

that the enzyme activity in nearly anhydrous media depends

(just like it does in conventional aqueous medium) on the

ionization state of the active-site residues. Thus the so-called

pH memory operates because, upon freeze-drying, side

chains of amino acids retain their ionized/un-ionized state.

Hence, the freeze-dried powders show enhanced activity

as a result of pH tuning. However, In aqueous solutions,counterions can freely move around in a solution. Because

they are not closely associated with opposite charges, their

identity does not affect the protonation state of the enzyme.

Thus, pH alone governs the protonation state. When a

biocatalyst is suspended in a low water organic solvent, the

situation is more complex. In this case, counterions are in

closer contact with the opposite charges on the enzyme

because of the lower dielectric constant of the medium.

Thus, protonation of ionizable groups on the enzyme will be

controlled by the type and availability of these ions as well

as hydrogen ions [68].

Several other ways of fixing the apparent pH of enzyme

powders have been described in the literature.

(a) Organic soluble buffers

Some pairs of acids and their salts are available which are

soluble in organic solvents. Blackwood et al. [69] used

triphenylacetic acid and its sodium salt to optimize the

catalytic performance of subtilisin in pentanone. Similarly,tri-iso-octylamine and its hydrochloride salt were used to

enhance the activity of lipase in pentanone [69]. For use in

more non-polar solvents, dendritic polybenzyl ether and its

sodium salt have been described [70]. A list of such soluble

buffers has been provided elsewhere [68].

(b) Solid-state buffers

These pairs remain insoluble in the reaction medium, but

exchange H+ /counterions with the protein molecule. Each

buffer pair, of course, fixes only one value of the ionization

parameter. An example of this approach is the use of



8/13


lysine/lysine hydrochloride for control of pH + pCl (= log

[Cl]) in the case of subtilisin-catalysed reaction in hexane

and toluene [71].

Vakos et al. [72] have exploited pH memory effect

to develop selectivity for a chemical modification reaction.

It was shown that, when proteins were freeze-dried from

aqueous solutions at pH values between 6.0 and 7.0, the

reaction with iodomethane in a vacuum was restricted to

-amino groups. Reaction with 13C-labelled iodomethane

made tentative identification of N-terminal amino acids

by 13C-NMR possible. The approach is especially valuable

in checking whether a protein has a free or blocked N-

terminus. This innovative idea can be extended to many

other situations wherein it may be advantageous to limit

chemical modification with not-so-selective reagents to

a specific kind of side chain. Identification of active-site

residues of enzymes, fine tuning an affinity label reactionand chemical cross-linking with a right trade off between

stabilization versus activity loss are some possible potential

applications. As many monofunctional and bifunctional

chemical modification reagents are soluble in organic

solvents, the major constraint is that such reactions

have to be carried out with suspensions of freeze-dried

powders.

Use of excipients

In an early report, using 98 % (w/w) KCl as an excipient

during freeze-drying led to a subtilisin preparation with

4000-fold higher kcat in hexane [73]. Later, optimization led

to salt activation, resulting in a 20 000-fold increase inkcatfor

the same reaction [74]. Ru et al. [75] have used the Jones

Dole coefficient to study the effect of salt activation. This

parameter evaluates an additive for its influence over water

viscosity and thus can be related to preferential hydration

of the protein in its presence (see the earlier discussion of

preferential exclusion theory under Protein stability during

freezethawing). Salt activation has also been shown to

work for -chymotrypsin, thermolysin, lipase and penicillin

amidase [66]. An important application of salt activation,

cited in [66], is the thermolysin-catalysed regioselectiveacylation of taxol to form the adipic acid derivative, which

is 1700 times more soluble in water and thus overcomes

the problem of the low solubility of taxol in water in its

therapeutic uses. Similarly, salt-activated subtilisin-acylated

doxorubicin showed enhanced potency (in efficacy against

a breast-cancer cell line) as compared with doxorubicin. In

both cases, the inactivated freeze-dried enzyme showed no

significant activity.

An important and key work in this regard is from

Mattiassons group [76], since it describes somewhat

different results and conclusions. Their experience was that

with -chymotrypsin, . . . addition of potassium chloride

could not yield any improvement in activity for the pre-

paration lacking buffer species [76]. Another important

observation was that while the specific activity of freeze-

dried enzyme increased with increasing amounts of buffer

salts, it decreased with higher amounts of sodium phosphate

(above 35 mmol/g of protein). The authors noted the earlier

finding that sodium phosphate differs from other salts

because of its capacity to associate with water in distinct

structures [76]. The activating effect of sorbitol was also

enhanced when buffer salts were present. The effect of these

additives on the transesterification activity of immobilized

-chymotrypsin has also been described by these authors.

Recently, in our studies with -chymotrypsin [77] and

tannase [78], the effect of salt activation was found to be only

marginal. Additionally, data with a wider range of enzymes

are needed before any definite conclusions about the effectof some additives, especially KCl (during freeze-drying), can

be drawn.

The presence of urea during freeze-drying led to

substantially activated subtilisin and peroxidase preparations

[79]. The effect of urea may not be restricted to partial

unfolding of the enzyme before lyophilization, as mentioned

in [66]; a more likely explanation is that given by the authors

themselves: urea, by H-bonding, may reduce dehydration-

induced structural changes. This is confirmed by the fact

that no such effect was observed with another denaturant,

namely guanidinium chloride [79]. These authors have

provided valuable data on the comparative efficacy of salt

activation and urea activation. The activation of subtilisin in

hexane and tetrahydrofuran was lower for urea than for KCl

(126-fold versus 2103-fold in hexane; 45-fold versus 122-fold

in tetrahydrofuran), but higher for urea in acetone

(18-fold versus 12-fold).

Molecular imprinting or inhibitor-induced activation

is a somewhat different kind of approach wherein the

substrate analogue of an enzyme is present during freeze-

drying. Russel and Klibanov [80] have shown that the

presence of a competitive inhibitor during freeze-drying

increased the activity of subtilisin 100-fold. Later, similar

results were reported for peroxidases, chloroperoxidases

and myoglobin [66]. In the case of lipase from the yeastCandida antarctica, enantioselectivity of the enzyme was

enhanced by molecular imprinting [81]. Rich et al. [82]

have used molecular imprinting with paclitaxel-2-adipic acid

and salt activation to enhance the activity of thermolysin

110-fold, the two activating approaches being found to be

additive in nature.

FTIR spectroscopy has shown that the use of a crown

ether resulted in subtilisin having a secondary structure in

dioxane very similar to that of the native enzyme [66].

Crown ethers as lyoprotectants also protect the enzyme by

forming non-covalent complexes through the-NH2groups



9/13


of lysine residues of enzymes [66]. Santos et al. [83] believe

that crown ethers act as molecular imprinters. As crown-

ether-activated enzymes show enhanced activity in polar

solvents, whereas salt-activated enzymes show increased

activity in non-polar solvents, it is possible that both ad-

ditives activate enzymes by different mechanisms. Cyclo-

dextrins, the glucose oligomers with cavities of diameter

0.470.83 nm, are known to stabilize proteins in aqueous

solutions [84]. Freeze-drying of subtilisin along with methyl-

-cyclodextrin increased the transesterification activity of

the enzyme powder 164-fold in tetrahydrofuran [85].

The effect of cyclodextrin was seen only in hydrophilic

solvents, but not in hydrophobic ones like toluene and

octane. Additionally, this chiral additive enhanced the

enantioselectivity of subtilisin and Candida rugosa lipase.

Interestingly, the enantioselectivity of lipase was opposite

to that of subtilisin.

Alternatives to the use of freeze-dried

enzyme powders

Over the years, workers have generally switched over to

using enzymes in forms other than simple freeze-dried

powders. These are reported to give better catalytic activ-

ities in nearly anhydrous media.

Immobilization Both adsorption and covalent coupling

[86,87] have been tried. Solgel methods have also been

reported [88].

Cross-linked enzyme crystals [89] Altus Biologics, Inc.

(Cambridge, MA, U.S.A.) has a patented technology for

creating CLECsTM (cross-linked enzyme crystals). These

are essentially microcrystals of enzymes which have been

cross-linked. These are reported to be extremely robust

preparations and show high stability at high temperatures as

well as in organic solvents.

Soluble enzymes [90,91] It was found as early as the 1980s

that enzymes coupled to poly(ethylene glycol) become

soluble in nearly anhydrous solvents. Otamari et al. [92]

showed that some complexes of enzymes with polymers

are also soluble in organic solvents.

Surfactant-modified enzymes Modification of enzymes, es-

pecially lipases, with surfactants gives a preparation that

solubilizes well, disperses evenly in organic solvents and

gives higher inter-esterification activity with 1,3-positional

specificity. The lipase from Pseudomonas sp. was modified

with a surfactant solution of sorbitan monostearate [93]

to hydrolyse palm oil most efficiently. Isono et al. [94]

have described the esterification activity of sorbitan

monostearate-modified lipase from Rhizopus japonicus for

wax-ester synthesis.

Enzyme precipitates This avoids the step of freeze-drying.

The enzyme is precipitated from its solution in aqueous

buffer by adding an organic solvent [95]. Various solvents,

such as 2-methylpropan-2-ol (t-butyl alcohol), propan-2-

ol (isopropanol), ethanol or 2-methylbutan-2-ol (t-amyl

alcohol) have been recommended [96]. The excess water

is removed by repeated suspension and precipitation of

the enzyme solution with the co-solvent. It should be

added that precipitation with organic solvents has been a

classical practice in enzymology for the concentration and

fractionation of enzymes. Further details on this approach

may be found elsewhere [9799]. On the basis of this

approach, Moore et al. have described propanol-rinsed

enzyme preparations (PREPs) [100].

Post-freeze-drying storage stability

Constantino et al. [101] have provided examples of how

freeze-dried proteins can become denatured in the presence

of moisture. BSA, freeze-dried from pH 7.3 and stored

at 37 C, with a water content of approx. 40 g/100 g

of protein, underwent aggregation (resulting in diminished

solubility). Aggregation was via thioldisulphide interchange.

Similar results were obtained for recombinant human

albumin. In both cases, freeze-drying from acidic solutions

(which ensures that SH groups are not ionized) pre-

vented aggregation (and the resultant solubility problem).

Previously, Constantino et al. [102] reported seven excipi-

ents which could prevent moisture-induced aggregation

and demonstrated that their stabilizing potency could be

correlated with their water-sorbing power. In the case

of freeze-dried human insulin at 50 C and 96% relative

humidity, aggregation via -elimination occurred, followed

by thiol-catalysed disulphide exchange. Again, lowering the

pH of the solution before freeze-drying reduced both

-elimination and subsequent thioldisulphide exchange.

The third example discussed is that of tetanus toxoid,

which illustrates yet another mechanism of deterioration

of a freeze-dried protein during storage. Tetanus toxoid

formed aggregates, caused by covalent non-disulphide

bonds. These bonds were found to be between Schiff-baseintermediates (generated from formaldehyde molecules

present in the toxoid preparation) and the side chains

of lysine, tyrosine and histidine residues in the toxoid

molecule. The preventive strategy was succinylation of

toxoid or reduction of labile formaldehyde linkages

with cyanoborohydride. However, in their previous work

[103], the authors reported that freeze-drying produced a

reversible reduction in -helix content (with concomitant

increase in-sheet structure). Thus, probably, freeze-drying

per se caused only reversible changes in the secondary

structure, and aggregation via non-disulphide covalent bonds



10/13


was a subsequent phenomenon, i.e. during storage and in the

presence of moisture.

Arakawa et al. [2] have also discussed some examples

concerning storage stability. In interleukin-2, the non-

essential residue Cys125 was changed to alanine or serine

and led to enhanced storage stability by preventing

aggregation. Similarly, in the case of fibroblast growth

factor, mutation of two solvent-exposed cysteine residues

improved the shelf-life of the freeze-dried protein. Free

cysteine residues (e.g. in granulocyte colony-simulating

factor), if buried and unavailable to react with other cysteine

residues, do not cause a problem during storage. In general,

inactivation mechanisms at this stage are identical with

those which have been known to operate in solution.

Hence all the strategies that work in stabilizing proteins

in solution should work at the storage stage, which

is, after all, dominated by moisture-triggered structuralchanges.

Finally, it is necessary to recall the relevance ofTg to

storage stability, which has been briefly referred to above

(see the section Basic process). It is believed that proteins

are more stable when stored at temperatures below their Tg[9]. It has, however, been pointed out that, although storage

below Tg is advantageous, it does not guarantee stability.

This is because molecular mobility still exists below this

temperature [8]. It is also necessary to realize that even a

trace of moisture may bring the Tg to storage temperature

and cause deactivation. It has been suggested that, as a

rule of thumb, the dried products should be stored at least

50 C below the Tg [8]. This, however, may quite often be

impractical.

Freeze-drying: miscellaneous issues

Use of freeze-drying is not restricted to proteins alone.

It has become a widely used process in many areas of

chemistry, pharmacy and biology. Brown et al. [104] have

examined the stability of retinal,-tocopherol, trans-lyopene

and trans--carotene in freeze-dried serum. Chongprasert

et al. [105] showed that the crystal forms of the drug

pentamidine isethionate present after freeze-drying dependon the initial solution concentration and the thermal history

of the sample before freezing. It is pointed out that

seemingly subtle differences in processing conditions can

have a significant impact on the critical quality attributes

of freeze-dried products. Earlier, van Winden et al. [106]

had looked at the stability of liposomes during freeze-

drying and concluded that slow freezing resulted in a

marked retention of encapsulated fluorescent molecule

after freeze-drying and rehydration, as compared with rapid

freezing. Thus quick freezing affects the integrity/stability of

liposomes.

The freeze-dried plasma standard for calibrating the

fibrinogen assay was introduced in 1992 by The National

Institute for Biological Standards and Control (NIBSC,

South Mimms, Potters Bar, Herts., U.K.). A recent study

examined the influence of freeze-drying on the clot-

ting properties of fibrinogen in plasma [107]. The authors

concluded, . . . findings support the use of a freeze-dried

international plasma calibrator, also in fibrinogen assays

based on measurements of clotting rate. In epidemio-

logical studies, however, when comparing minor differences

in fibrinogen concentrations, the influence of freeze-drying

on the clotting rate of calibrations plasma should be

taken into account [107]. In general, there is obviously

a need to be cautious when using freeze-dried proteins

in analytical methods. Haddeland et al. [108] have de-

scribed an interesting example of how freeze-drying can

sometimes give rise to a novel biological property. It wasfound that freeze-dried fibrinogen stimulated tissue-type-

plasminogen-activator-catalysed plasminogen activator. The

conformational changes during freeze-drying presumably

exposed stimulatory sites which were inaccessible in the

native fibrinogen molecule.

Conclusions

DePaz et al. [109] pointed out that, in some cases, storage

stability of some peptides/protein (in solid/powder form)

is even less than in aqueous solutions. Thus a greater

understanding of the freeze-drying process may result

in a rational, and hopefully more stable, formulation for

enzymes and pharmaceuticals. If freeze-drying damages the

protein conformation (and catalytic activity), why did it

escape the notice of biochemists? In other words, why was

freeze-drying perceived as a gentle way of concentrating

proteins? The answer to this question lies in the fact

that enzymology (and our expectations from enzymes!)

have undergone subtle changes over the years. Earlier,

biochemists were using proteins/enzymes only in aqueous

buffers. As structural changes during freeze-drying are

mostly reversible during rehydration, the deleterious effects

of freeze-drying escaped notice. Today, enzymes are usedin nearly anhydrous organic solvents [4], reverse micelles

[110], solvent-free systems [111] and the frozen state

[112]. Also, production processes are expected to result in

preparations with reproducible biological activity, especially

in the case of pharmaceutical proteins and protein-based

vaccines. Driven by a balance sheet, it is also desirable

to have as highly active a catalyst as possible during

process optimization. Hence the current focus on this

overlooked process, which was used rather empirically.

It has been pointed out that one of the limiting factors

in gene therapy is the instability of viruses [86]. For



11/13


example, retroviruses get inactivated during freeze-drying

rather easily. It is likely that our understanding of freeze-

drying can ultimately be extended to these more complex

nucleoproteins. This optimistic note is not without basis.

Trehalose, which protects protein structure during thermal

stress and drying, does have a protective function on more

complex structures. Trehalose is widely found in nature

among diverse species which are capable of withstanding

conditions where they are exposed to conditions of low

water activity and/or thermal stress, e.g. bacteria, yeast,

mushroom, nematodes, shrimp and desert plants [113].

Thus freeze-drying (lyophilization) may turn out to be

a Cinderella among the techniques which enzymologists

have been using over the years. The fascination with this

technique, although of relatively recent origin, is already

paying rich dividends in many process designs [114,115].

Acknowledgments

The financial assistance from the Department of Science

and Technology, Department of Biotechnology, Council

for Scientific and Industrial Research (both Extramural

Division and Technology Mission on Oilseeds, Pulses and

Maize) and National Agricultural Technology Project (Indian

Council for Agricultural Research), all Government of India

organizations, is gratefully acknowledged.

References

1 Maa, Y. F. and Prestelski, S. J. (2000) Curr. Pharm. Biotechnol.

1, 283302

2 Arakawa, T., Prestelski, S. J., Kenney, W. C. and Carpenter, J. F.

(2001) Adv. Drug Deliv. Rev.46, 307326

3 Mazzobre, M. F., Longinotti, M. P., Corti, H. R. and Buera, M. P.

(2001) Cryobiology43, 199210

4 Gupta, M. N. (1992) Eur. J. Biochem.203, 2532

5 Gupta, M. N. (Ed.) (2000) Methods in Nonaqueous Enzym-

ology, Birkhauser Verlag, Basel

6 Vulfson, E. N., Halling, P. J. and Holland, H. L. (eds.) (2001)Enzymes in Nonaqueous Solvents: Methods and Protocols,

Humana Press, Totowa, NJ

7 Fagain, C. O. (1996)in Protein Purification Protocols(Doonan,

S., ed.), pp. 323327, Humana Press, Totowa, NJ

8 Craig, D. Q. M., Royall, P. G., Kett, V. L. and Hopton, M. L.

(1999) Int. J. Pharm.179, 179207

9 Sun, W. Q., Davidson, P. and Chan, H. S. O. (1998) Biochim.

Biophys. Acta145, 245254

10 Heller, M. C ., Carpenter, J. F. and Randolph, T. W. (1999) Arch.

Biochem. Biophys.363, 191201

11 Hatley, R. H. (1992) Dev. Biol. Stand.74, 105119

12 Yoshioka, S., Aso, Y., Kojima, S. and Tanimoto, T. (2000) Chem.

Pharm. Bull.48, 283285

13 Mazzobre, M. F. and Buera, M. D. P. (1999) Biochim. Biophys.

Acta1473, 337344

14 Rupley, J. A., Gratton, E. and Careri, G. (1983) Trends Biochem.

Sci.8, 1822

15 Poole, P. L. and Finney, J. L. (1983) Int. J. Biol. Macromol. 5,

308310

16 Falconi, M., Brunelli, M., Pesce, A., Ferrario, M., Bolognesi, M.

and Desideri, A. (2003) Proteins: Struct. Funct. Genet. 51,

607615

17 Klibanov, A. M. and Ahern, T. J. (1987) in Protein Engineering

(Oxender, D. L. and Fox, C. F., eds.), pp. 213218, Alan R. Liss,

New York

18 Gray, C. G. (1993) in Thermostability of Enzymes (Gupta,

M. N., ed.), pp. 124143, Springer Verlag, Heidelberg

19 Rosevear, A. (1988) in Molecular Biology and Biotechnology(Walker, J. M. and Gingold, E. B., eds.), pp. 255284, Royal

Society of Chemistry, London

20 Crestfield, A. M., Stein, W. H. and Moore, S. (1962) Arch.

Biochem. Biophys., Suppl. 1, 217222

21 Fruchter, R. G. and Crestfield, A. M. (1965) J. Biol. Chem.240,

38753882

22 Mozhaev, V. V. (1993) Trends Biotechnol.11, 8895

23 Gupta, M. N. (ed.) (1993) Thermostability of Enzymes,

Springer Verlag, Heidelberg

24 Schmid, R. D. (1979) Adv. Biochem. Eng.12, 41117

25 Patro, S. Y., Freund, E. and Chang, B. S. (2002) Biotechnol.

Annu. Rev.8, 5584

26 Webb, S. D., Golledge, S. L., Cleland, J. L., Carpenter, J. F. and

Randolph, T. W. (2002) J. Pharm. Sci. 91, 14741487

27 Gupta, M. N. (1991) Biotechnol. Appl. Biochem.14, 111

28 Millqvist-Fureby, A., Malmsten, M. and Bergenstahl, B. (1999)

Int. J. Pharm.191, 103114

29 Eckhardt, B. M., Oeswein, J. Q. and Bewley, T. A. (1991) Pharm.

Res.8, 13601364

30 Hsu, C. C., Nguyen, H. M., Yeung, D. A., Brooks, D. A., Koe,

G. S., Bewley, T. A. and Pearlman, R. (1995) Pharm. Res. 12 ,

6977

31 Sarciaux,J. M., Mansour, S., Hageman, M. J. and Nail,S. L. (1999)

J. Pharm. Sci. 88, 13541361

32 Cao, E., Chen, Y., Cui, Z. and Foster, P. R. (2003) Biotechnol.Bioeng.82, 684690

33 Crowe, J. H., Carpenter, J. F., Crowe, L. M. and Anchordoguy,

T. J. (1990) Cryobiology27, 219231

34 Arakawa, T. and Timasheff, S. N. (1983) Arch. Biochem.

Biophys. 224, 169177

35 Timasheff, S. N. (2002) Proc. Natl. Acad. Sci. U.S.A. 99,

97219726

36 Carpenter, J. F. and Crowe, J. H. (1988) Cryobiology 25,

459470

37 Crowe, J. H. and Carpenter, J. F. (1989) Biochemistry 28,

39163922



12/13


38 Griebenow, K. and Klibanov, A. M. (1995) Proc. Natl. Acad.

Sci. U.S.A.92, 1096910976

39 Lippert, K. and Galinski, E. A. (1992) Appl. Microbiol.

Biotechnol.37, 6165

40 Prestrelski, S. J., Tedeschi, N., Arakawa, T. and Carpenter, J. F.

(1993) Biophys. J.65, 661667

41 Carpenter, J. F., Prestrelski, S. J. and Arakawa, T. (1993) Arch.


42 Prestrelski, S. J., Arakawa, T. and Carpenter, J. F. (1993) Arch.


43 Strambini, G. B. and Gabellieri, E. (1996) Biophys. J. 70,

971976

44 Carrasquillo, K. G., Sanchez, C. and Griebnow, K. (2000)

Biotechnol. Appl. Biochem. 31, 4153

45 Jiang, S. and Nail, S. L. (1998) Eur. J. Pharm. Biopharm. 45,

249257

46 Anchordoguy, T. J. and Carpenter, J. F. (1996) Arch. Biochem.

Biophys. 332, 231238

47 Yoshioka, S., Aso, S. and Kojima, Y. (1999) Pharm. Res. 16,

135140

48 Dastoli, F. R., Musto, N. A. and Price, S. (1966) Arch. Biochem.

Biophys. 115, 4447

49 Dastoli, F. R. and Price, S. (1967) Arch. Biochem. Biophys.118,

163165

50 Dastoli, F. R. and Price, S. (1967) Arch. Biochem. Biophys.122,

289299

51 Zaks, A. and Klibanov, A. M. (1985) Proc. Natl. Acad. Sci. U.S.A.

82, 31923196

52 Inada, Y., Takahashi, K., Yoshimoto, T., Ajima, A., Matsushima,A. and Saito, Y. (1986) Trends Biotechnol. 4, 190194

53 Zaks, A. and Klibanov, A. M. (1988) J. Biol. Chem. 263,

31943201

54 Zaks, A. and Klibanov, A. M. (1988) J. Biol. Chem. 263,

80178021

55 Reslow, M., Adlercreutz, P. and Mattiasson, B. (1988) Eur. J.

Biochem. 177, 313318

56 Halling, P. J. (1987) Biocatalysis1, 109115

57 Griebenow, K., Vidal, M., Baez, C., Santos, A. M. and Barletta,

G. (2001) J. Am. Chem. Soc. 123, 53805381

58 Klibanov, A. M. (1986) Chemtech16, 354359

59 Khmelnitsky, Y. L., Mozhaev, V. V., Belova, A. B. , Sergeeva, M. V.and Martinek, K. (1991) Eur. J. Biochem. 198, 3141

60 Gupta, M. N., Batra, R., Tyagi, R. and Sharma, A. (1997)

Biotechnol. Prog.13, 284288

61 Laane, C., Boeren, S., Vos, K. and Veeger, C. (1987) Biotechnol.

Bioeng. 30, 8087

62 Dordick, J. S. (1992) Biotechnol. Prog.8, 259267

63 Mattiasson, B. and Adlercreutz, P. (1991) Trends Biotechnol.9,

394398

64 Triantafyllou, A. O., Adlercreutz, P. and Mattiasson, B. (1993)

Biotechnol. Appl. Biochem. 17, 167179

65 Klibanov, A. M. (1997) Trends Biotechnol.15, 97101

66 Lee, M.-Y. and Dordick, J. S. (2002) Curr. Opin. Biotechnol.13,

376384

67 Zacharis, E., Moore, B. D. and Halling, P. J. (1999) Proc. Natl.

Acad. Sci. U.S.A.96, 12011205

68 Partridge, J., Harper, N., Moore, B. D. and Halling, P. J. (2001)

in Enzymes in Nonaqueous Solvents: Methods and Protocols

(Vulfson, E. N., Halling P. J. and Holland, H. L., eds.), pp.

227234, Humana Press, Totowa, NJ

69 Blackwood, A. D., Curran, L. J., Moore, B. D. and Halling, P. J.

(1994) Biochim. Biophys. Acta 1206, 161165

70 Dolman, M., Halling, P. J. and Moore, B. D. (1997) Biotechnol.

Bioeng.55, 278282

71 Zacharis, E., Moore, B. D. and Halling, P. J. (1997) J. Am. Chem.

Soc. 119, 1239612397

72 Vakos, H. T., Kaplan, H., Black, B., Dawson, B. and Hefford,

M. A. (2000) J. Protein Chem. 19, 231237

73 Khmelnitsky, Y. L., Welch, S. H., Clark, D. S. and Dordick, J. S.(1994) J. Am. Chem. Soc.116, 26472648

74 Ru, M. T., Dordick, J. S., Reimer, J. A. and Clark, D. S. (1999)

Biotechnol. Bioeng.63, 233241

75 Ru, M. T., Hirokane, S. Y., Lo, A. S., Dordick, J. S., Reimer, J. A.

and Clark, D. S. (2000) J. Am. Chem. Soc. 122, 15651571

76 Triantafyllou, A. O., Wehtje, E., Adlercreutz, P. and Mattiasson,

B. (1997) Biotechnol. Bioeng. 54, 6776

77 Roy, I. and Gupta, M. N. (2003) Tetrahedron59, 54315436

78 Sharma, S. and Gupta, M. N. (2003) Bioorg. Med. Chem. Lett.

13, 395397

79 Guo, Y. and Clark, D. S. (2001) Biochim. Biophys. Acta1546,

406411

80 Russel, A. J. and Klibanov, A. M. (1988) J. Biol. Chem. 263,

1162411626

81 Secundo, F., Carrea, G., Soregaroli, C., Varinelli, D. and

Morrone, R. (2001) Biotechnol. Bioeng. 73, 157163

82 Rich, J. O., Mozhaev, V. V., Dordick, J. S., Clark, D. S.

and Khmelnitsky, Y. L. (2002) J. Am. Chem. Soc. 124,

52445245

83 Santos, A. M., Vidal, M., Pacheco, Y., Frontera, J., Baez, C.,

Ornellas, O., Barletta, G. and Griebnow, K. (2001) Biotechnol.

Bioeng.74, 295308

84 Cooper, A. (1992) J. Am. Chem. Soc.114, 92089209

85 Montanez-Clemente, I., Alvira, E., Macias, M., Ferrer, A.,

Fonceca, M., Rodriguez, J., Gonzalez, A. and Barletta, G. (2002)Biotechnol. Bioeng.78, 5359

86 Tanaka, A. and Kawamoto T. (1991) in Protein Immobilization:

Fundamentals and Applications (Taylor, R. F., ed.), pp. 183208,

Marcel Dekker Inc., New York

87 Bosley, J. and Peilow, A. (2000) in Methods in Nonaqueous

Enzymology (Gupta, M. N., ed.), pp. 5169, Birkhauser Verlag,

Basel

88 Chen, J.-P. and Lin, W.-S. (2003) Enzyme Microb. Technol.32,

801811

89 Margolin, A. L. and Navia, M. A. (2001) Angew. Chem. Int. Ed.

Engl.40, 22042222



13/13


90 Takahashi, K., Kodera, Y., Yoshimoto, T., Ajima, A., Matsushima,

A. and Inada, Y. (1985) Biochem. Biophys. Res. Commun.131,

532536

91 Matsushima, A., Kodera, Y., Takahashi, K., Saito, Y. and Inada, Y.

(1986) Biotechnol. Lett. 8, 7378

92 Otamari, M., Adlercreutz, P. and Mattiasson, B. (1992)

Biocatalysis 6, 291305

93 Isono, Y., Nabetani, H. and Nakajima, M. (1996) Bioprocess

Eng.15, 133137

94 Isono, Y., Nabetani, H. and Nakajima, M. (1995) J. Am. Oil

Chem. Soc.72, 887890

95 Partridge, J., Halling, P. J. and Moore, B. D. (1998) Chem.

Commun.841, 842

96 Partridge, J., Hutcheon, G. A., Moore, B. D. and Halling, P. J.

(1996) J. Am. Chem. Soc.118, 1287312877

97 Harris, E. L. V. (1989) in Protein Purification Methods: A

Practical Approach (Harris, E. L. V. and Angal, S., eds.),pp. 125174, IRL Press, Oxford

98 Englard, S. and Seifter, S. (1990) Methods Enzymol. 182,

285300

99 Doonan, S. (1996) in Protein Purification Protocols (Doonan,

S., ed.), pp. 135144, Humana Press, Totowa, NJ

100 Moore, B. D., Partridge, J. and Halling, P. J. (2001) in Enzymes

in Nonaqueous Solvents: Methods and Protocols (Vulfson,

E. N., Halling P. J. and Holland, H. L., eds.), pp. 97104, Humana

Press, Totowa, NJ

101 Constantino, H. R., Schwendeman, S. P., Langer, R. and Klibanov,

A. M. (1998) Biochemistry (Moscow) 63, 357363

102 Constantino, H. R., Langer, R. and Klibanov, A. M. (1995)

Biotechnology (N.Y.)13, 493496

103 Constantino, H. R., Schwendeman, S. P., Griebenow, K.,

Langer, R. and Klibanov, A. M. (1996) J. Pharm. Sci. 85,

12901293

104 Brown, T. J., Duewer, D. L., Kline, M. C . and Sharpless, K. E.

(1998) Clin. Chim. Acta 276, 7587

105 Chongprasert, S., Griesser, U. J., Bottorff, A. T., Williams,

N. A., Byrn, S. R. and Nail, S. L. (1998) J. Pharm. Sci. 87,

11551160

106 van Winden, E. C ., Zhang, W. and Crommelin, D. J. (1997)

Pharm. Res.14, 11511160

107 Jensen, T., Halvorsen, S., Godal, H. C. and Skjonsberg, O. H.

(2002) Thromb. Res.105, 499502

108 Haddeland, U., Sletten, K., Bennick, A. and Brosstad, F. (1994)

Blood Coagul. Fibrinolysis5, 575581

109 DePaz, R. A., Dale, D. A., Barnett, C. C., Carpenter, J. F.,

Gaertner, A. L. and Randolph, T. W. (2002) Enzyme Microb.

Technol. 31, 765774

110 Melo, E. P., Costa, S. M., Cabral, J. M., Fojan, P. and Petersen,

S. B. (2003) Chem. Phys. Lipids124, 3747

111 Won, K., Jeong, J. C. and Lee, S. B. (2002) Biotechnol. Bioeng.

79, 795803

112 Kamat, S. V., Beckmen, E. J. and Russel, A. J. (1995) Crit. Rev.

Biotechnol.15, 4171

113 Murray, B. S. and Liang, H.-J. (2000) Langmuir16, 60616063

114 Hancock, B. C. and Dalton, C. R. (1999) Pharm. Dev. Technol.4, 125131

115 Pendas, J., Moreira, T., Guerra, O., Pena, B. R. and Fernandez,

J. A. (2001) Cryo-Letters 22, 512

116 Dabulis, K. and Klibanov, A. M. (1993) Biotechnol. Bioeng.41,

566571

117 Ru, M. T., Wu, K. C., Lindsay, J. P., Dordick, J. S., Reimer, J. A.

and Clark, D. S. (2001) Biotechnol. Bioeng.75, 187196

118 Dai, L. and Klibanov, A. M. (1999) Proc. Natl. Acad. Sci. U.S.A.

96, 94759478

119 Altreuter, D. H., Dordick, J. S. and Clark, D. S. (2002) J. Am.

Chem. Soc.124, 18711876

120 Persson, M., Mladenoska, I., Wehtje, E. and Adlercreutz, P.

(2002) Enzyme Microb. Technol. 31, 833841

121 Rich, J. O. and Khmelnitsky, Y. L. (2001) Biotechnol. Bioeng.72,

374377

122 Lindsay, J. P., Clark, D. S. and Dordick, J. S. (2002) Enzyme

Microb. Technol.31, 193197

Received 28 July 2003/15 October 2003; accepted 11 December 2003

Published as Immediate Publication 11 December 2003,

DOI 10.1042/BA20030133


Freeze-drying of Proteins

Documents

Transcript of Freeze-drying of Proteins